my-model-Bertin-Area

This model is a fine-tuned version of bertin-project/bertin-roberta-base-spanish on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.1711
• Accuracy: 0.4903
• F1: 0.4767
• Precision: 0.5366
• Recall: 0.4903

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 30
• eval_batch_size: 5
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
1.8349	1.0	25	1.7703	0.3032	0.2608	0.2991	0.3032
1.7709	2.0	50	1.7153	0.3355	0.2347	0.2079	0.3355
1.7315	3.0	75	1.6515	0.3613	0.2934	0.2937	0.3613
1.596	4.0	100	1.6332	0.3871	0.3405	0.3123	0.3871
1.3388	5.0	125	1.6449	0.4065	0.3298	0.2900	0.4065
1.0783	6.0	150	1.5722	0.4581	0.3797	0.3760	0.4581
0.8663	7.0	175	1.6593	0.3935	0.3563	0.3490	0.3935
0.6118	8.0	200	1.9177	0.4581	0.4481	0.4658	0.4581
0.4206	9.0	225	2.2944	0.4065	0.3920	0.4286	0.4065
0.3375	10.0	250	2.2870	0.4387	0.4359	0.4889	0.4387
0.2334	11.0	275	2.4912	0.4065	0.4015	0.4546	0.4065
0.1618	12.0	300	2.5429	0.4710	0.4499	0.5204	0.4710
0.1238	13.0	325	2.7109	0.4710	0.4458	0.5135	0.4710
0.0906	14.0	350	2.8377	0.4774	0.4594	0.5092	0.4774
0.071	15.0	375	3.0123	0.4839	0.4656	0.5461	0.4839
0.0498	16.0	400	3.0204	0.4710	0.4517	0.4959	0.4710
0.0416	17.0	425	3.0939	0.4839	0.4622	0.5107	0.4839
0.0281	18.0	450	3.0979	0.4903	0.4793	0.5281	0.4903
0.0226	19.0	475	3.1622	0.4839	0.4708	0.5202	0.4839
0.0185	20.0	500	3.1711	0.4903	0.4767	0.5366	0.4903

Framework versions

• Transformers 4.41.0
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1